DESCRIPTION (APPLICATION ABSTRACT): Patients in the hospital, young and old, are sicker than ever and constitute a group highly vulnerable to medical error. While dozens of studies have documented rates of medication administration errors averaging 10 percent or more, almost nothing is known about how working conditions contribute to medication administration error. Still less is known about the factors that promote actions by nurses to intercept potential errors and prevent adverse drug effects. The purpose of this study is to describe the relationships between working conditions that affect nurses and the safety and quality of care they provide, with a focus on medication safety. The emphasis of this study is organizational variables that can be affected administratively, rather than individual nurse characteristics. Specific aims of this study are to (1) develop a computer- based system for collection of observational data on the full scope of nurses' functions in medication administration; (2) develop a measure of safety climate at the level of the nursing unit; (3) develop and test a predictive model of medication administration safety that encompasses nurses' working conditions (e.g., physical environment, safety climate, automation, and staffing levels), as well as nurse workload and actions (e.g., shift length, patient assignment, adherence to standards, and actions that prevent adverse drug effects); (4) evaluate the relationship between medication safety and other indicators of quality of care; and (5) disseminate evidence on working conditions associated with medication safety. The computerized data collection tool will allow an observer to efficiently capture medication administration errors, adherence to drug administration standards, nurses' error reduction activities, and environmental and systems factors impinging on safe administration of medications. The computerized observation tool and the safety climate tool will be developed, validated, and tested in the first year in two local hospitals. The tools will then be used to collect data in 12 hospitals across the country (chosen to provide variation in regional and hospital characteristics). Observing four nurses on each of five units in each hospital will provide data on 240 nurses, at least 6,000 doses, 600 to 1,200 errors, and 2,400 sets of doses (all the drugs administered at one time). The safety climate survey will be administered to nurses on each of the 60 units in the study, and existing data will be gathered regarding the nurse staffing levels, rates of other adverse incidents occurring on these units. Using data from these sources, hierarchical linear analysis will be conducted to test a predictive model of medication safety.